Continuous flow synthesis of pyridinium salts accelerated by multi-objective Bayesian optimization with active learning†

Nuclear magnetic resonance (NMR) spectroscopy is a valuable and complementary tool in environmental research, but it is underutilized due to the cost, size, and maintenance requirements of standard "high-field" NMR spectrometers. "Low-field" NMR spectrometers are a financially and physically accessible alternative, but their lower sensitivity and increased spectral overlap limit the analysis of heterogeneous environmental/biological media, especially with fast-relaxing samples that produce broad, low-intensity spectra. This study therefore investigates the potential of the steady-state free precession (SSFP) experiment to enhance signal-to-noise ratios (SNRs) of fast-relaxing, complex samples at both high- and low-field. SSFP works by obtaining steady-state transverse signal using a train of equally spaced radiofrequency pulses with the same flip angle and a time between pulses less than the transverse relaxation time, allowing for thousands of scans to be summed in a short time period. Here, 13C-SSFP is applied to samples of varying complexity (egg white, dissolved organic matter, and crude oil) at low-field and at high-field for testing and comparison. The potential of in vivo SSFP NMR is additionally investigated by applying 31P-SSFP to live Eisenia fetida at high-field. In some samples, SSFP increased 13C SNR by over 2000% at both high-field and low-field compared to standard 13C NMR and enabled detection of peaks that were not observable by standard 13C NMR. Ultimately, SSFP holds great potential for improving analysis of fast-relaxing, complex samples, which could in turn make low-field NMR spectroscopy a more effective tool not only in environmental/biological research but also in numerous other disciplines.

A comparative evaluation of low-field and high-field NMR untargeted analysis: Authentication of virgin coconut oil adulterated with refined coconut oil as a case study

In multi-objective Bayesian optimization, an infill criterion is an important part, as it is the indicator to evaluate how much good a new set of solutions is, compared to a Pareto-front approximation set. This paper presents a deterministic algorithm for computing the Expected R2 Indicator for bi-objective problems and studies its use as an infill criterion in Bayesian Global Optimization. The R2-Indicator was introduced in 1998 by M. Hansen and A. Jaszkiewicz for performance assessment in multi-objective optimization and is more recently also used in indicator-based multi-criterion evolutionary algorithms (IBEAs). In Bayesian Global Optimization, we propose the Expected R2-indicator Improvement (ER2I) as an infill criterion. It is defined as the expected decrease of the R2 indicator by a point that is sampled from a predictive Gaussian distribution. The ER2I can also be used as a pre-selection criterion in surrogate-assisted IBEAs. It provides an alternative to the Expected Hypervolume-Indicator Improvement (EHVI) that requires a reference point, bounding the Pareto front from above. In contrast, the ER2I works with a utopian reference point that bounds the Pareto front from below. In addition, the ER2I supports preference modelling with utility functions and its computation time grows only linearly with the number of considered weight combinations. It is straightforward to approximate the ER2I by Monte Carlo Integration, but so far a deterministic algorithm to solve the non-linear integral remained unknown. We outline a deterministic algorithm for the computation of the bi-objective ER2I with Chebychev utility functions. Moreover, we study monotonicity properties of the ER2I w.r.t. parameters of the predictive distribution and numerical simulations demonstrate fast convergence to Pareto fronts of different shapes and the ability of the ER2I Bayesian optimization to fill gaps in the Pareto front approximation.

The aim of this study was to discriminate the authenticity of perilla oils distributed in Korea using their 1H nuclear magnetic resonance (NMR) spectra acquired by a 43 MHz low-field benchtop NMR spectrometer. Significant differences existed in the integration values of all 6 peaks found in the spectrum between authentic and adulterated perilla oil samples. The integration values of 4 peaks that signify the methylene protons present in all fatty acids (FA) and allylic or olefinic protons present in all unsaturated FA were the best variables for establishing perilla oil authenticity. The procedure for applying the range of variables found in authentic perilla oil samples correctly discriminated between the samples of perilla oils with soybean oils added at concentrations of ≥ 6 vol%. The results demonstrated that this NMR procedure is a possible cost-effective alternative to the high-field 1H NMR method for discriminating the authenticity of perilla oils.

The measurement of self-diffusion coefficients using pulsed-field gradient (PFG) nuclear magnetic resonance (NMR) spectroscopy is a well-established method. Recently, benchtop NMR spectrometers with gradient coils have also been used, which greatly simplify these measurements. However, a disadvantage of benchtop NMR spectrometers is the lower resolution of the acquired NMR signals compared to high-field NMR spectrometers, which requires sophisticated analysis methods. In this work, we use a recently developed quantum mechanical (QM) model-based approach for the estimation of self-diffusion coefficients from complex benchtop NMR data. With the knowledge of the species present in the mixture, signatures for each species are created and adjusted to the measured NMR signal. With this model-based approach, the self-diffusion coefficients of all species in the mixtures were estimated with a discrepancy of less than 2 % compared to self-diffusion coefficients estimated from high-field NMR data sets of the same mixtures. These results suggest benchtop NMR is a reliable tool for quantitative analysis of self-diffusion coefficients, even in complex mixtures.

A low-field nuclear magnetic resonance spectrometer for non-invasive monitoring of human finger blood glucose fluctuations was developed. Saline solution and blood serum samples with different glucose concentrations were first detected by the spectrometer and it has been found that there was a high-linear correlation between the glucose concentration and the transverse relaxation time. Then, the spectrometer was employed to noninvasively measure a finger from each of the several volunteers. The experiment results showed that the transverse relaxation time of the human finger increases with human blood glucose concentration. In conclusion, the human finger nuclear magnetic resonance spectrometer could be a potential tool to noninvasive monitoring of human body’s blood glucose fluctuations in the future.

We propose a surrogate-assisted reference vector adaptation (SRVA) method to\nsolve expensive multi- and many-objective optimization problems with various\nPareto front shapes. SRVA is coupled with a multi-objective Bayesian\noptimization (MBO) algorithm using reference vectors for scalarization of\nobjective functions. The Kriging surrogate models for MBO is used to estimate\nthe Pareto front shape and generate adaptive reference vectors uniformly\ndistributed on the estimated Pareto front. We combine SRVA with expected\nimprovement of penalty-based boundary intersection as an infill criterion for\nMBO. The proposed algorithm is compared with two other MBO algorithms by\napplying them to benchmark problems with various Pareto front shapes.\nExperimental results show that the proposed algorithm outperforms the other two\nin the problems whose objective functions are reasonably approximated by the\nKriging models. SRVA improves diversity of non-dominated solutions for these\nproblems with continuous, discontinuous, and degenerated Pareto fronts.\nBesides, the proposed algorithm obtains much better solutions from early stages\nof optimization especially in many-objective problems.\n

Paul Christian Lauterbur

A hardware system for a laboratory-built desktop nuclear magnetic resonance (NMR) spectrometer was constructed with a module design based on single-phase detection architecture. This spectrometer features an operating frequency of DC to 100 MHz, RF pulse power of 0.7244 mW to 1.2589 W, minimum detectable signal of 1.5136 × 10−10 mW, and receiver bandwidth of 20 kHz. Verification of the functional feasibility of NMR detection was implemented using a 30-mL custom probe in a 0.23-T magnetic field. A baseband free induction decay was also acquired through the application of an RF pulse at 1 W for 10 μs. The observed initial amplitude and frequency of the MR signal are 1.47 V and 1.25 kHz, respectively. Meanwhile, the effective transverse relaxation time () for this configuration declines to no more than 15 ms because no shimming treatments were taken. The proposed desktop NMR spectrometer is suitable for building a hand-made NMR spectrometer in the laboratory, and enhances opportunities for novel and exciting NMR experiments.

Field homogeneity of a superconducting magnet is a critical factor for nuclear magnetic resonance (NMR) spectrometers. For REBCO high-temperature superconducting (HTS) coil fabrication, layer-wound is considered a better option than double-pancake (DP) winding, in terms of field homogeneity. Nonetheless, DP winding is preferred to layer-wound owing to the ease of manufacturing and the difficulty in making long REBCO tapes. However, DP winding has an inevitable crossover turn between the top and bottom layers. For magnets with an HTS insert used in NMR spectrometers, the crossover turn could be critical for the field homogeneity. In this paper, the magnetic field generated by the crossover turn is derived based on the Biot-Savart law, and the effect of crossover turns in stacked DPs on the field homogeneity of an NMR-class HTS magnet is analyzed.

Bayesian optimization (BO) methods solve problems with several black-box objectives and constraints.Each black-box is expensive to evaluate and lacks a closed-form.They use a model of each black-box to guide the search for the problem's solution.Sometimes, however, the black-boxes may be evaluated at different fidelity levels.A lower fidelity is simply a cheap proxy of the corresponding black-box.Thus, lower fidelities that correlate with the actual black-box can be used to reduce the optimization cost.We propose Joint Entropy Search for Multi-Fidelity and Multi-objective Bayesian Optimization with Constraints (MF-JESMOC), a BO method for solving the aforementioned problems.It chooses the next point and fidelity level at which to evaluate the black-boxes as the one that is expected to reduce the most the joint entropy of the Pareto set and the Pareto front, normalized by the fidelity's cost.Deep Gaussian processes are used to model each black-box and dependencies between fidelities.In our experiments, MF-JESMOC outperforms other state-of-the-art methods for multi-objective BO with constraints and different fidelity levels.

Wounds care and management is one of the most basic needs in the medical setting. Burn wounds, trauma wounds, pressure ulcers and bedsores are just some of the many types of wounds that need to be treated quickly and efficiently. Take for instance pressure ulcers, there at least 3 million reported cases of chronic ulcers and bedsores reported in the U.S. each year alone. Each ulcer can takes weeks to months to fully heal, leading to extended hospital stays and decreased quality of life in patients. Current treatment seeks to protect and keep the wound clean, manage cellular exudates, and in general reduce wound area. This is done by packing antibiotic gels into the wound bed, which have a bolus drug release profile, and covering the area with commercially available bandages. There are many studies documenting bolus drug release profiles and their effects on wounds, but what is distinctly absent are studies which investigate the effects of continuous drug delivery on wound healing. Here it is hypothesized that applying a continuous flow of fresh drugs into the wound bed will speed up the often lengthy wound healing process. To examine this hypothesis, human umbilical vein endothelial cells (HUVEC) and adult human dermal fibroblasts (HDFa) cell culture models were used to mimic the wound healing process. Results were characterized by the rate of cell migration and proliferation (wound closure) into a simulated wound channel created in the middle of nearly confluent HUVEC and HDFa cell culture models. Both cell types were exposed to static or continuous flow conditions with and without drug infused media. For the HUVEC cells testing indicates a significant difference between the average rate of closure for flow versus no flow conditions (0.0628 versus 0.0232 percent closure per minute, p = 0.00165). For the HDFa cells testing indicates a significant difference between the average rate of closure for flow versus no flow conditions (0.0595 versus 0.0392 percent closure per minute, p = 0.01606). The research suggests that positive growth rate occurs in HDF and HUVEC when continuous low flow conditions are applied to the cells.

In this study, changes in the transverse relaxation time (T2), signal intensity (A2), and images of ‘Jinxiu’ yellow peach fruits were investigated using low-field nuclear magnetic resonance (LF-NMR) and proton magnetic resonance imaging (1H-MRI). The water content, (determined using centrifuging technology), fruit firmness, and ethylene production were evaluated at 4±1°C and 25±2°C. The results showed a signal peak (A2) and T2 at 500–1000 ms. On days 4–6, T2 changed from 796 ms to 861 ms, and decreased with longer time at 25°C. Water content also decreased and high adjusted determination coefficient (adj-R2) values with water content were found for T2 (R2 = 0.95) and A2 (R2 = 0.99), indicating a high degree of fit for the prediction equations. A2 decreased and T2 shortened after prolonged storage at 4°C and high adj-R2 values were found between the water content and T2 (R2 = 0.93) and A2 (R2 = 0.97).Ethylene production and firmness were the same as for T2, A2, and internal breakdown at 25°C for 4–6 days, and 4°C for 16–21 days. Compared with storage at 25°C, low temperature storage inhibited water loss and also prohibited internal breakdown. T2 and A2 can be used to evaluate the storage characteristics of yellow peach fruits, and the internal breakdown of pulp tissue could be clearly observed by 1H-MRI, which is a function of LF-NMR. 1H-MRI can provide high-resolution images showing the internal structure of intact peaches. Thus, LF-NMR is a new technology that could be an effective and simpler method for monitoring and predicting water content, as well as detecting quality issues during peaches storage.

Coffee brings many health benefits due to its chemical constituents. Based on this information, it is essential to know the main chemical compounds from coffee granules; the intermolecular interaction among the coffees compounds and the molecular components homogeneity. In this study six types of roasted commercial coffee were evaluated by solid state nuclear magnetic resonance (NMR), employing carbon-13 (13C) and hydrogen (1H) nucleus. Carbon-13 was analyzed applying high field NMR techniques, such as: magic angle spinning (MAS); magic angle spinning with cross-polariza- tion (CPMAS) and magic angle spinning with cross-polarization and dipolar dephasing (CPMASDD). The hydrogen was evaluated via relaxation times. Proton spin-lattice relaxation time in the rotating frame was deter-mined through the carbon-13 decay, during the variable contact-time experiment, using high field NMR. Proton spin-lattice relaxation time was determined through the inversion-recovery pulse sequence, using low field NMR. Considering all NMR results, it was concluded that the major coffee compounds are: a) triacilglycerides, which constitute the mobile region in the granule coffee and b) Carbohydrates such as: polysaccharides and fibers that belong to the rigid domain. These constituents belong to different molecular mobility domain, although they have strong intermolecular interactions due to the granule organization.

New proposal of mechanical reinforcement structures to annular REBaCuO bulk magnet for compact and cryogen-free NMR spectrometer